Decomposition strategy for mathematical programming models of solvent design
Date of Completion
January 2005
Keywords
Engineering, Chemical
Degree
Ph.D.
Abstract
This dissertation presents a novel decomposition methodology for the solution of mathematical programming models for computer-aided molecular design (CAMD) problems. The framework consists of formulating the CAMD problem as an MINLP model, solution of the MINLP model through the decomposition approach and analysis of the results through verification steps. ^ This dissertation focuses on design of solvents and solvent mixtures for crystallization processes through the above-mentioned framework. Solvent design framework is developed for two types of crystallization processes namely cooling crystallization and drowning out crystallization. A number of solvent characteristics such as solubility, potential recovery, flammability, toxicity, viscosity, boiling point and melting point are considered within this framework. In addition to the above-mentioned properties, the effect of solvents on the crystal morphology, which is a qualitative property, is also considered within the design framework. Group contribution models are used for the estimation of properties. Case studies related to design of solvents and/or anti-solvents for Ibuprofen an important pharmaceutical compound is solved. ^ In order to verify the performance of the solvents with respect to crystal morphology wet lab crystallization experiments are carried out. The morphology and the structure of crystals are analyzed through scanning electron microscope (SEM) and powder x-ray diffraction (p-xrd) techniques.^
Recommended Citation
Thirugnanam Karunanithi, Arunprakash, "Decomposition strategy for mathematical programming models of solvent design" (2005). Doctoral Dissertations. AAI3195557.
https://digitalcommons.lib.uconn.edu/dissertations/AAI3195557